Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
Message-ID: <83a807111f17c9e2db4e9f12a7ec1437c11d0f55.camel@ndufresne.ca>
Subject: Re: [PATCH v4 07/15] media: mtk-vcodec: vdec: add media device if
 using stateless api
From:   Nicolas Dufresne <nicolas@ndufresne.ca>
To:     Alexandre Courbot <acourbot@chromium.org>,
        Hans Verkuil <hverkuil-cisco@xs4all.nl>
Cc:     Tiffany Lin <tiffany.lin@mediatek.com>,
        Andrew-CT Chen <andrew-ct.chen@mediatek.com>,
        Rob Herring <robh+dt@kernel.org>,
        Yunfei Dong <yunfei.dong@mediatek.com>,
        Mauro Carvalho Chehab <mchehab@kernel.org>,
        Linux Media Mailing List <linux-media@vger.kernel.org>,
        LKML <linux-kernel@vger.kernel.org>,
        "moderated list:ARM/Mediatek SoC support" 
        <linux-mediatek@lists.infradead.org>
Date:   Thu, 13 May 2021 11:07:40 -0400
In-Reply-To: <CAPBb6MUKtxCS=JbtBmvwwEovrN8NCtLmMGcTkozo_gnDMRsqHw@mail.gmail.com>
References: <20210427111526.1772293-1-acourbot@chromium.org>
         <20210427111526.1772293-8-acourbot@chromium.org>
         <faa5553f-2ea5-27a5-7f85-e1418d2c7df1@xs4all.nl>
         <CAPBb6MUKtxCS=JbtBmvwwEovrN8NCtLmMGcTkozo_gnDMRsqHw@mail.gmail.com>
Content-Type: text/plain; charset="UTF-8"
User-Agent: Evolution 3.40.1 (3.40.1-1.fc34) 
MIME-Version: 1.0
Content-Transfer-Encoding: 8bit
Precedence: bulk

Le jeudi 13 mai 2021 à 17:05 +0900, Alexandre Courbot a écrit :
> Hi Hans, thanks for the review!
> 
> On Thu, Apr 29, 2021 at 4:28 PM Hans Verkuil <hverkuil-cisco@xs4all.nl> wrote:
> > 
> > On 27/04/2021 13:15, Alexandre Courbot wrote:
> > > From: Yunfei Dong <yunfei.dong@mediatek.com>
> > > 
> > > The stateless API requires a media device for issuing requests. Add one
> > > if we are being instantiated as a stateless decoder.
> > 
> > Why for the stateless decoder only? Why not create one for all?
> > 
> > It's not a blocker, but I would recommend looking at this.
> 
> Would there be any use in creating a media device for a stateful
> decoder that does not need to use requests?

The only thing I can think of is classification. In GStreamer support for
stateless decoders, I use the topology to classify what is a decoder by walking
the toplogy and making sure it's what I expect, and that there is no other
branches or functions that I may not support. This makes it more strict, so less
likely to confuse driver function.

Note that v4l2-compliance just use the same old heuristic we have used for
stateful, which is to check that one side have only RAW format, and the other
side have only encoded formats. It works too, it's just less strict.

> 
> Cheers,
> Alex.